Regular Paper

International Journal of Control, Automation and Systems 2018; 16(2): 749-758

Published online March 1, 2018

https://doi.org/10.1007/s12555-017-0231-y

© The International Journal of Control, Automation, and Systems

RBF Neural Network Sliding Mode Consensus of Multiagent Systems with Unknown Dynamical Model of Leader-follower Agents

Amin Sharafian, Vahid Bagheri, and Weidong Zhang*

Shanghai Jiao Tong University

Abstract

This paper proposed a new methodology to cover the problem of consensus of multiagent systems with sliding mode control based on Radial Basis Function (RBF) neural network. First, neural network adopted to distinguish the uncertainties of the leader and follower agents then a sliding mode tracking controller is applied to force the follower agents to follow the leader’s time-varying states trajectory with the consensus error as small as possible. As the RBF neural network is adopted to approximate the uncertainties, the results can only achieve local consensus. Different from past literature, total error of consensus protocol is considering for sliding surface therefore the local stability of the whole multiagent system is provided meanwhile RBF neural network overcome the problem of unmodeled leader/follower agent dynamics. The weights of the neural networks updated adaptively directly commensurate with consensus error. The point is, there is absolutely no need to have information about dynamical model of the system. The merits of the proposed approach are consisting of consensus protocol robustness, fast error convergence to zero, and local stability of the closed loop multiagent system which is proved by Lyapunov direct method. The simulation results show promising performance of the proposed method on a chaotic system."

Keywords Consensus, multiagent, RBF neural network, nonlinear systems, sliding mode.

Article

Regular Paper

International Journal of Control, Automation and Systems 2018; 16(2): 749-758

Published online April 1, 2018 https://doi.org/10.1007/s12555-017-0231-y

Copyright © The International Journal of Control, Automation, and Systems.

RBF Neural Network Sliding Mode Consensus of Multiagent Systems with Unknown Dynamical Model of Leader-follower Agents

Amin Sharafian, Vahid Bagheri, and Weidong Zhang*

Shanghai Jiao Tong University

Abstract

This paper proposed a new methodology to cover the problem of consensus of multiagent systems with sliding mode control based on Radial Basis Function (RBF) neural network. First, neural network adopted to distinguish the uncertainties of the leader and follower agents then a sliding mode tracking controller is applied to force the follower agents to follow the leader’s time-varying states trajectory with the consensus error as small as possible. As the RBF neural network is adopted to approximate the uncertainties, the results can only achieve local consensus. Different from past literature, total error of consensus protocol is considering for sliding surface therefore the local stability of the whole multiagent system is provided meanwhile RBF neural network overcome the problem of unmodeled leader/follower agent dynamics. The weights of the neural networks updated adaptively directly commensurate with consensus error. The point is, there is absolutely no need to have information about dynamical model of the system. The merits of the proposed approach are consisting of consensus protocol robustness, fast error convergence to zero, and local stability of the closed loop multiagent system which is proved by Lyapunov direct method. The simulation results show promising performance of the proposed method on a chaotic system."

Keywords: Consensus, multiagent, RBF neural network, nonlinear systems, sliding mode.

IJCAS
July 2024

Vol. 22, No. 7, pp. 2055~2340

Stats or Metrics

Share this article on

  • line

Related articles in IJCAS

IJCAS

eISSN 2005-4092
pISSN 1598-6446